INDEX

Explanations

political figures saying

The neuron detects proper personal names—mentions of specific people (e.g. Donald Trump, Barack Obama, Jeremy Corbyn, Lewis Hamilton).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

-1.91

of

-1.77

".

-1.70

!!

-1.70

!.

-1.69

 zarówno

-1.68

eração

-1.68

-1.67

."

-1.65

.,

-1.63

POSITIVE LOGITS

itize

1.39

 nastro

1.38

ਔ

1.37

ufe

1.31

tissimo

1.28

 bronzo

1.27

ᾐ

1.27

nhà

1.27

alors

1.26

apropri

1.26

Activations Density 0.049%